A Research On Fusion Multi-Sensor Indoor Positioning Algorithm Based On Geomagnetic Field And UWB

In recent years,Location-based Services have gradually penetrated into every aspect of life with the change of people's lifestyle.In outdoor environment,GPS provides users with efficient and accurate positioning as a mature positioning system.It becomes an urgent problem to provide users with accurate indoor positioning services with the increase of people's indoor activities.However,due to building occlusion or signal interference,the positioning accuracy of GPS is drastically reduced or even can not apply to positioning services.Based on above requirements,types of indoor positioning technologies were proposed,such as Wi-Fi positioning technology,Geomagnetic field positioning technology,Inertial sensor positioning technology and so on.The existing positioning technologies generally have the following problems.Firstly,most positioning technologies rely on environment configuration.For example,Wi-Fi positioning requires a large number of access points.Therefore,these methods are costly and the usage scenarios are greatly limited.Secondly,Most of the positioning technologies only provide the absolute positioning of a single object,the relative positioning of two or more targets is less studied.Assuming that the position information can be presented to each other through indoor positioning system,users will find the companions quickly and easily.Therefore,it is of great practical significance to develop a low-cost relative positioning technology without relying on external anchor points.Base on systematic research on indoor positioning technologies,in this paper firstly,a fusion multi-sensors geomagnetic indoor positioning algorithm based on improved particle filter is proposed to improve high-precision indoor positioning for a single target.Secondly,an indoor relative positioning algorithm which based on UWB ranging system and multisensors is proposed.The main work includes the following aspects:1.The implementation principle of the pedestrian dead reckoning method based on inertial sensors is deeply studied.Firstly,the peak detection method is used to detect the step number,and then the off-line training is used to obtain the step size estimation model.Aiming at the problem that the direction estimation error is large in the traditional PDR method,a multi-sensor direction estimation algorithm is proposed to improve the measurement accuracy.2.The principle of geomagnetic positioning technology based on traditional particle filter is studied.Firstly,the off-line magnetic fingerprint map is established,and then the principle and existing problems of traditional particle filter algorithm are analyzed.The RANSAC algorithm is used to determine initial position to assist in local particle initialization in the process of particle initialization,meanwhile,the improved PDR method algorithm is used to update the particle state in the process of particle state update.The experimental results show that the accuracy of the fusion multi-sensor geomagnetic positioning algorithm based on improved particle filter performs better.3.Under the premise of not relying on external anchor points,a multi-sensor indoor relative positioning algorithm based on ultra-wideband ranging is proposed aming at the lack of research on the relative positioning.For the ranging part of the algorithm,the ultra-wideband ranging technology is introduced,and the DWM1000 ultra-wideband wireless ranging system and its ranging principle are introduced in detail.The superiority of the centimeter level ranging accuracy of the system is verified.In addition,the relative positioning algorithm of the two points is extend to multi-points relative positioning,which enlarges the applicable scenarios of the proposed algorithm.The simulation results show that the proposed algorithm does not rely on external anchor points,and has low cost,simple implementation and high positioning accuracy.